The Challenge of Diagnosing Fanconi Anemia in Patients with Bone Marrow Failure (BMF): A Study in 82 BMF Patients.

Abstract

The Bone Marrow Failure (BMF) syndromes comprise a variety of distinct acquired or inherited clinical entities. Early distinction between syndromes has clear implications in the disease management and outcome. Fanconi anemia (FA), the most frequent cause of inherited BMF, is usually associated with congenital abnormalities, progressive cytopenia, chromosome fragility, and cancer susceptibility. However, due to a high clinical variability and/or the potential emergence of revertant hematopoietic cells (somatic mosaicism), identifying patients with FA or ruling out this diagnosis can be challenging. If undiagnosed, FA patients who initially present with bone marrow failure will die of toxicity after standard-dose conditioning regimens for HSCT. In this study, we evaluated FA diagnosis in patients with BMF but no clear initial evidence of FA, using of a combination of classical and innovative tests in blood and fibroblasts. A cohort of 82 patients with BMF and no strong clinical evidence of FA was analysed (patients with a clear FA diagnosis were not included). Based on the likelihood of an underlying inherited condition associated with the BMF, we classified patients in 3 groups: those likely to have idiopathic aplastic anemia (IAA) [n=38, group 1], those likely to have a constitutional condition other than FA [n=26, group 2], and those likely to have IAA but who had isolated clinical findings which could also be present in FA [n=18, group 3]. Chromosome breakage test and analysis of the FA/BRCA pathway by FANCD2 immunoblot were performed in PBL in all patients [n= 82]. To overcome potential somatic mosaicism, skin primary fibroblasts were analysed [n= 52]. Also, to rule out FA/BRCA downstream groups, we developed a new flow cytometry test based on MMC-sensitivity in fibroblasts. In total, 6 patients with FA were identified: 1/38 in group 1 (aplastic anemia at 10 yo, no positive clinical findings), 2/26 in group 2 (one with MDS at 48 yo with precocious menopause and vocal cord neoplasia at 38 yo; one with hypoplastic MDS at 50 yo), and 3/18 in group 3 (one with short stature and aplastic anemia at 37 yo; one with MDS and borderline physical abnormalities at 26 yo; and one with a single cafe-au-lait spot and aplastic anemia at 10 yo). Chromosomal breakage tests in PBL were sufficient to diagnose 4 of these FA patients (further classified as FA core using FANCD2 immunoblot). Additional fibroblast analyses were necessary to identify 2 more FA patients (both with complete somatic mosaicism) and, importantly, to definitely exclude FA diagnosis in other patients. In conclusion, underdiagnosing FA is rare if careful history and physical exam are done together with chromosome breakage test in PBL. However, in clinical situations where the suspicion of FA persists despite negative breakage tests, then fibroblasts should be tested. Because no cases of FA were found among patients with IAA and negative breakage tests in PBL, we suggest that FA screening can be limited to this technique in this population (group 1). The strategy here presented allowed us to identify a few unexpected FA cases in a cohort of BMF patients, and importantly, to definitely rule out FA in others, with clear clinical impact for patients who undergo HSCT.